The overall aim of this study is to test the "matching hypothesis" that alcohol treatment effectiveness can be increased by assigning clients with certain characteristics to particular treatments. The present application proposes to continue work initiated and conducted over the past five years. The specific aims of this study are: to tests primary and secondary a priori matching hypotheses over the course of 15 months of follow-up; to conduct psychometric and other analyses of patient, treatment process, and outcome variables to provide the basis for testing the matching hypotheses; to examine alternative analytic strategies and variables for testing matching; and to determine the extent to which matching effects persist over a three year period following treatment completion. Data sets collected from the outpatient (N=954) and aftercare (N=774) arms of the Project MATCH clinical trial will be analyzed to achieve these aims. Clients have been randomly assigned to one of three treatments, Twelve Step, Cognitive Behavioral, or Motivational Enhancement. They were assessed prior to treatment on baseline and matching variables, and were then followed for 15 months, at 3-month intervals, to measure their alcohol consumption and other dimensions of outcome. Clients in the outpatient arm of the study will be contacted and assessed at 39 months after treatment initiation. In Year 08, these data will be analyzed to test longer term matching effects. It is anticipated that the study results will impact on alcohol treatment research and delivery for the next decade. The Project MATCH Steering Committee, following discussions with NlAAA Staff, has chosen to submit a generic proposal across each of the CRUs, reflecting the common protocol and cooperative process that has guided the project across its first five years of operation. However, each CRU has provided site-specific information in the budget, budget justifications, listing of key personnel, consultant/consortium agreements, and human subjects sections of their respective proposals.